[AFS3-std] AFS Standardization Proposal

[Simon Wilkinson]

Sorry, I've been a bit busy with work, and have let moving this  
forwards slide. I believe that we have (so far) identified 3 issues,  
which I'l discuss in turn below. It would be good to reach some kind  
of consensus on the portions of these that are outstanding.

*) Copyright

Jeff has suggested some copyright boilerplate language to include at  
the head of all standardisation documents - there have been no issues  
raised with this, and so I intend to update the document to include  
this language unless anyone shouts.

*) Defining the Electorate

There have been a number of proposals for changes to the way in which  
the electorate is defined. There seem to be two separate issues - the  
first is whether we place some kind of eligibility hurdle in the way,  
and the second is how we handle 'dead' email addresses, so that the  
total size of the electorate doesn't grow without bound (which is  
important in order to handle the recall case).

Personally, I believe that establishing any kind of competence hurdle  
is going to be extremely difficult to manage. I'd be interested in  
proposals of exactly how such a hurdle could be defined without  
introducing a significant level of subjectivity to the electoral  
process. Without a competence definition that avoids the need to make  
subjective decisions, my personal view is that we can't introduce an  
eligibility requirement.

Managing dead email addresses seems best performed through mailman's  
standard mechanisms. Deleting all accounts which don't reply to  
email, or bounce email on a certain data runs the risk of  
disenfranchising users due to technical problems outside their control.

*) The home of the registrar (and the standards list)

There are two options here - either leaving things as they are at  
grand.central.org, or moving them to systems which are hosted by the  
OpenAFS foundation, in whatever form that takes. It's very difficult  
to judge this at present, without knowing what the foundation can  
provide. In my opinion we should remove the language about where  
these functions are hosted from the document at present, and discuss  
further when the foundation has come into being.

Comments and arguments greatly appreciated ...

Simon.